H100-hosted Polish LLM

Slayer v49 Qwen3.5-27B

OpenAI-compatible API running on a dedicated H100 behind the Fabryka Vultr bridge. Built for Polish style, instruction following, and measurable usage.

Use the API
Total requests
7
Total tokens
0
24h requests
7
Avg latency
0.038s

Model

slayer-v49-qwen3.5-27b
LoRA over Qwen3.5-27B, served by vLLM.

Bridge

H100 on lem -> reverse SSH tunnel -> Vultr nginx -> public HTTPS.

Stats

Usage is counted per request, token usage, latency, and key hash. Public aggregates are on /stats.